Skip to main content

Full text of "Johnson & Johnson : building an infrastructure to support global operations"

See other formats


II I II Hill lllll II Will llll llll llll II Hill Hill II 111! Illl II ! II 

3 9080 01439 0741 








Jeanne W. Ross 

September 199S 

CISR WP No. 283 

Sloan WP No. 38S1 

©1995 Massachusetts Institute of Technology 

Center for Information Systems Research 

Sloan School of Management 

Massachusetts Institute of Technology 

77 Massachusetts Avenue, E40-193 

Cambridge, MA 02139-4307 

SEP i o igg ? 



Johnson & Johnson: 
Building an Infrastructure to Support Global Operations 

Jeanne W. Ross 
MIT Sloan School of Management 


Johnson & Johnson, Inc. has over 100 years' experience operating in decentralized management practices, 
but its customers are increasingly demanding that J&J present "a single face." This means that J&J must 
coordinate sales, distribution, financial, and marketing information across its 160 operating companies. 
This case reviews possible strategies for developing an information technology infrastructure that would 
enable J&J to be more responsive to its customers' demands. The case cites both technical and 
organizational challenges inherent in implementing a more centralized infrastructure. 

This is one in a series of case studies developed for MIT's executive education course on "Managing the 
IT Network for Global Competitiveness. " The case study and accompanying teaching note are intended 
to describe and analyze one company 's experiences in building and managing its IT infrastructure. 

Center for Information Systems Research 
Massachusetts Institute of Technology 
Sloan School of Management 

Johnson & Johnson: 
Building an Infrastructure to Support Global Operations 


On January 1, 1995, Johnson and Johnson (J&J) established J&J Health Care Systems (HCS) whose 
mission was to provide J&J products to large managed care and provider organizations. HCS was a 1 ,200 
person company representing the J&J U.S. pharmaceutical, diagnostic, medical/surgical and consumer 
companies to customers like HMOs, integrated delivery systems and hospital organizations. At the same 
time, it was a center of excellence defining the needs of this new breed of customer to the J&J operating 
companies. HCS was a response to the changing health care industry. Dennis Longstreet, Chairman of 
J&J HCS, explained: 

The industry itself is reshaping and it's brought on by the desire for the payor to 
focus on the economics of health care. What's happened is that stand-alone 
hospitals and physicians, who had been our primary customers for health care 
products, are no longer the sole decision-makers. It's become an integrated delivery 
system, where the doctor and the hospital and the payor and insurance company are 
all becoming more connected to focus on delivering cost-effective quality health 

J&J HCS was the second company that Johnson & Johnson had created to market products of existing 
companies to large customers. Johnson & Johnson's Customer Support Center was created in 1992 to sell 
J&J consumer products to large U.S. retailers like Wal-Mart and KMart. Jim Litts, President of the 
Customer Support Center, noted that his efforts to work closely with six different operating companies 
represented a counter-cultural approach to work at J&J: 

J&J has over 100 years of history authorizing operating companies to manage all 
business facets to maximize their brands' P&Ls. Today, we are learning how 
difficult it is to break those paradigms and work together to leverage the strength 
of Johnson & Johnson with larger retail customers. 

While HCS and the Customer Support Center were different from J&J's usual independent operating 
company model, Longstreet and Litts felt they were representative of how J&J would operate in the 
future. The two executives noted that the inter-company cooperation and coordination demanded by this 
organizational model had significant implications for J&J's culture and for the amounts and kinds of 
information that would be communicated and shared across J&J operating companies. 

This case was prepared by Jeanne Ross at the MTT Center for Information Systems Research. It may be freely 
distributed to students in not-for-profit educational institutions. The author would like to thank Mike Vitale, Jack 
Rockart and Debra Hofman for their helpful comments and the many Johnson & Johnson employees who 
contributed insights, cooperation and support. 

Copyright ®I995 Massachusetts Institute of Technology. 


Johnson & Johnson, with 1994 sales of over $15 billion, was the world's largest manufacturer of health 
care products Founded in 1886 as the first manufacturer of sterile dressings, the company had nearly 
doubled in size since 1987 and typically depended for one-third of its revenues on products that had been 
introduced within the prior five years. J&J sold products ranging from baby shampoo to treatments for 
leukemia and from disposable contact lenses to stents that could be inserted in arteries to improve the 
results of balloon angioplasty. In 1995, J&J had approximately 80,000 employees in about 160 operating 
companies, with markets in over 150 countries world-wide. (See Appendix A for a representative list of 

Johnson & Johnson had a long history of managing its operating companies as independent businesses. 
Corporate executives, dating back to Robert Wood Johnson in the 1930s, embraced operating company 
autonomy as a path to increased flexibility, accountability and creativity. Independent analysts also 
credited the decentralized J&J management approach as largely responsible for the corporation's 
consistently strong financial performance 1 . The independence of the individual units, however, meant 
that J&J employees tended to view themselves as employees of a particular J&J operating company rather 
than of the corporation. There was rarely any movement of employees between operating companies, and 
operating company executives were compensated based on the performance of their company, not the 
corporation as a whole. Consequently, J&J companies often regarded one another more as competitors 
than as members of the same team. 

By the early 1990s, top executives noted that J&J's autonomous operating companies were not well- 
positioned to service customers who were trying to limit the number of their vendor interactions. Each 
operating company had its own marketing and sales arm that worked directly with its customers. Matthew 
Martin, Vice President of Information Services for HCS, explained one consequence of this arrangement: 

Prior to the formation of Health Care Systems, each of the operating companies 
had a national accounts representative. Johnson & Johnson did not focus as a single 
corporation on its top customers. We could have up to 18 representatives from 
different J&J Companies calling on a customer. Eventually, we listened when the 
customer said, "Time out! Why can't Johnson & Johnson send me one person to 
deal with to negotiate a contract. It's more efficient for me and it must be for you 

Over time, corporate management introduced a variety of structures to mitigate the limitations of the 
decentralized management approach and increase inter-company cooperation. For example, the operating 
companies were organized into three groups: Consumer, Pharmaceutical, and Professional, and the 
chairman of each group was given responsibility for identifying opportunities for leveraging services and 
expertise across companies in each of these markets. Franchise managers were assigned responsibility for 
coordinating cross-company sales of a family of products, such as the baby care products of operating 
companies like Johnson & Johnson Consumer in the U.S., Johnson & Johnson France, and Johnson & 
Johnson Pacific Pty. Ltd. in Australia. Finally, the introduction of HCS and the Customer Support Center 

'See Tanouye, Elyse, "Johnson & Johnson Stays Fit by Shuffling Its Mix of Businesses," Wall Street 
Journal, December 22, 1992, p. Al, and Weber, Joseph, "A Big Company That Works," Business Week, May 
4, 1992, pp. 124-132. 

represented radically new ways to organize work at J&J. These companies focused on working across 
U.S. companies to address the needs of U.S. customers, but they could eventually be expanded or similar 
organizations could be introduced in other countries. 

When the operating companies had been completely autonomous, they had little need to share data. Most 
information flowed between a company and its customers, while financial data flowed from the company 
to corporate headquarters. Consequently, information systems, computing platforms, and data definitions 
grew up in J&J around individual company needs. As headquarters attempted to work across companies, 
management found that existing information systems and information system structures did little to 
facilitate those efforts. IS and business executives felt a need to build an information infrastructure that 
would respond to J&J's changing customer demands. 

J&J's Information Technology Infrastructure 

Consistent with J&J's decentralized approach to management, most information technology management 
responsibility was distributed to the operating companies. Each company typically had an independent 
information system unit responsible for systems planning, development, operations and maintenance. 
Operating company IT units also hired all their own IT staffs and were responsible for their compensation 
and professional development. While historically there had been little cross-company coordination among 
IT professionals, Group IT Vice Presidents were appointed in 1993 and IT directors from the operating 
companies had dotted line reporting responsibility to them. (See the organization chart for the Corporate 
Office of Information Technology in Appendix B.) 

While most infrastructure support, such as LAN management, help desk, desktop support, and local 
computer and telecommunications operations was provided by operating company IS departments, a small 
centralized IT function was based in the corporate data center in New Jersey. Called Networking and 
Computing Services (NCS), this centralized unit was responsible for the data center, but its primary 
responsibility was for managing J&J's global network and providing mainframe computing services for 
all J&J businesses in the U.S. 

J&J's global network was a traditional multiplexed Tl network providing telephone and dial-up data links 
between J&J headquarters, operating companies, and related facilities throughout the world. The fifty 
persons in the Corporate Network Services unit of NCS were responsible for contract negotiation and 
administration of telecommunications contracts, data network engineering and design, remote PBX and 
voicemail management, videoconferencing, and limited Internet support. NCS had not historically 
provided systems management or support for end-users and applications programmers, in part because 
the network environment was not conducive to centralized support. The operating companies had built 
a maze of subnetworks on a wide variety of computing platforms and Network Services did not have the 
network management tools, the breadth of expertise, or the charter to manage those subnetworks. 

While most of J&J's operating companies received network support directly from Corporate Network 
Services in New Jersey, European companies (Western and Eastern Europe, Middle East, and Africa) 
received support from a regional center in Belgium. The European regional center managed a router- 
based, single transport, primarily TCP/IP network from one central location. This network was a subset 
of J&J's global network and supported 100 European J&J locations with a backbone of over one hundred 
routers. More than a thousand servers were connected to the network and European Network Services 
staff managed the routers for all the local LANs to ensure that no one at a company site could configure 
a LAN in a manner that would jeopardize someone else. The tightly controlled nature of the network 

enabled a team of eleven J&J employees and six contractors to offer centralized support to European 

This team not onJy managed the physicaJ part of the network (telecom lines, routers, voice multiplexers) 
but had increasingly emphasized deploying enterprise network applications like e-mail, groupware, 
executive support systems, affiliate communication, and set-up. Jos DeSmedt, Director of European 
Network Services, noted some implications of centralized network management: 

The design and management [of the European Network] facilitates very tight LAN 
and WAN integration. Since there are no subnetworks for individual companies or 
franchises anymore, the Network management becomes much more critical. On the 
other hand, we can automate the management more uniformly over the region from 
this central location. 

The European Network Services unit had evolved from the Janssen Pharmaceutical IT unit, which 
serviced the largest operating company in Europe. Because many European operating companies were 
small, there were sometimes just a couple IT people addressing the needs of entire countries. Over time, 
they had purchased services from Janssen, which had resulted in many operating companies adopting 
Janssen standards for hardware and software. Thus, when European Network Services was formally 
established in July 1994, much of the service it provided had already been centralized. 

Although U.S. companies were typically larger and more self-sufficient with regard to their IT needs. 
Bob Chaput, Vice President of Networking and Computing Services, considered the European network 
a model for J&J. He anticipated developing additional regional network service centers in Asia and Latin 
America. More immediately, he intended to upgrade the services available from the corporate facility. 
He created a team in his unit to evaluate and support infrastructure applications as well as a team to 
develop new network services. (See the Network and Computing Services organization chart in Appendix 
C.) He noted, however, that for these teams to fully realize their potential, the Networking and 
Computing Services organization would have to take a more proactive role in defining networks within 
the operating companies: 

We know that [centralized network support] will work because we've been 
successful in Europe. The difference is the companies in the U.S. are bigger and 
stronger. They have more people and they fight harder and longer to retain control 
and independence. But the businesses' applications people generally are happy to 
have some stability in infrastructure applications like e-mail and Notes to have 
something they know works and something they know is supported 24x7. 

In early 1995 IT management identified four limitations with J&J's current infrastructure to help the 
company adapt to changing business conditions, particularly initiatives like HCS and the Customer 
Support Center. First, the amount of IS attention allocated to infrastructure management across the 
company was diluting the attention that could be focused on more strategic IT applications. Second, the 
lack of technology standards was inhibiting connectivity, aggravating attempts to service business needs, 
and costing too much to support. Third, the funding process for infrastructure projects was retarding 
efforts to build an enterprise-wide infrastructure. Finally, lack of data standards was impeding the 
meaningful exchange of data across companies. 

Allocation of IS human resources 

Bob Chaput estimated that 550 of J&J's approximately 1 ,500 IS professionals were engaged in supporting 
infrastructure technologies in the U.S. alone. He felt that centralizing functions such as telecommunica- 
tions support, help desk, desktop and local area network management, and computer operations could cut 
that number in half, even if most of the staff remained physically located in operating companies. His 
goal was to free up IT resources to work on higher business value projects through increased centraliza- 
tion of infrastructure responsibilities in order to gain economies of scale and eliminate redundant work. 

As a start, five major Professional Group companies in the U.S. had agreed to turn over responsibility 
for voice communications to Chaput's organization. In addition to its usual responsibility for working 
with vendors to design and install connections, Networking and Computing Services would have 
continuing management responsibility for telecommunications tasks such as voice mail and PBX moves, 
adds, and changes for the Professional Group companies. Warren Koster, Vice President of Information 
Technology for the Professional Group, noted that the companies expected centralized services to yield 
significant savings as well as some less tangible benefits: 

What we are driving to is leveraging the components of the infrastructure in 
Professional Group companies and driving costs out. At the same time, it's not just 
to drive out costs. It's to get people working on other projects that are more 
competitive and higher on the value chain and not worrying about the infrastructure 

Koster acknowledged that, despite the apparent efficiencies, there would be hesitancy to move towards 
shared services like this, because of concerns about potential personnel shifts and decreases in service 
levels. The Professional Group companies were preparing to centralize their distributed systems 
management, and some operating company IT directors expressed concern that this responsibility should 
remain local because it demanded more personalized service than telecommunications, which was viewed 
as a commodity service. But while some IT managers were anxious about increased centralization of 
infrastructure responsibilities, others were enthusiastic supporters. Carolyn McQuade, Vice President of 
Information Technology for the Consumer Group, wanted to leverage expertise: 

We need to extend centralized management control of the infrastructure down to 
the desktop level. The amount of time that we all spend debugging software like 
WordPerfect 6.0 is just ridiculous. We all load the same software, discover the 
same bugs and go through the same experiences as many times as there are 
companies. It's a shameful waste. Some organizations have more talented people 
in that area than others. We really could do a much better job of leveraging what 
we know. 

Establishing Information Technology Standards 

Not surprisingly, the autonomy of the IT units at the operating companies had led to great variety in the 
technologies they employed. On the hardware side, there was variation in technologies like routers and 
bridges and small wars between Macintosh and Windows computer users. On the software side, J&J had 
nine different email systems, frequent debates about desktop products, and a variety of network operating 
systems. Jan Fields, Director of Corporate Network Services, noted that enforcing a limited set of 
standards was key to enabling centralization of infrastructure services: 

You can't possibly build the skills for half a dozen different kinds of routers and 
bridges, for example. It's foolish to do that. Managing nonstandard equipment 
when you have a problem, trying to sectionalize and troubleshoot it, and get the 
correct vendors involved — all of those kinds of things add a tremendous amount 
of time to solving any kind of problem. 

The need to integrate different companies' systems and provide communication links for J&J HCS and 
the Customer Support Center highlighted the limitations of diversity in technologies. Networking and 
Computing Services had established standards but operating companies were not always quick to adopt 
them. Nonetheless, the Customer Support Center's Jim Lifts noted that he expected IT to establish 
standards and affiliated companies to conform to them: 

In my mind the IT community ought to come out and talk about the hardware and 
software to run this stuff. I think we ought to stop giving the operating companies 
votes. My point is, the software doesn't matter. Everybody will complain about it 
anyway. So let the IT guys make the economical, efficient choice, understanding, 
of course, what the user requirements are. 

Funding Infrastructure Investments 

Infrastructure development efforts by both Corporate and European Network Services were requisitioned 
by operating company management, who had to pay for whatever services they received. This charging 
mechanism sometimes acted as a deterrent to infrastructure investments. Bob Chaput provided an 

We'll have a franchise manager sitting in our Consumer business in New Jersey 
asking for manufacturing information from our plant in Singapore. Well, guess 
who's not connected? And so, the franchise manager will say, "Well, just go knock 
on their door and tell them you're here to install it." So we dutifully go out and 
knock on the door and say, "We're here to install your network connections at 
$1000 a month." And the local management says, "Time out, I don't have a 
thousand dollars a month." 

Jan Fields noted that individual operating companies did not always see the benefit of infrastructure 
investments. Start-up businesses, in particular, might feel that limited funds were better spent elsewhere: 

They can say, "We can dial up for email and if we get it a day later or two days 
later it doesn't matter." They want to put their money where it is going to impact 
their customer. They may say to us "For five different countries, I expect to pay 
a total of no more than $2000 a month." You can't deliver service for that. But if 
that's what it is worth to them, then that's what it is worth. 

Cross-company organizations like HCS and the Customer Support Center required that operating 
companies make changes for the good of J&J, even when the cost to the company seemed high. Funding 
processes that charged individual companies for infrastructure development could negatively impact 
investment levels. Chaput was trying to move discussions on infrastructure funding to higher 
organizational levels: 

When I go out into an operating company, I do a proposal, you react to it and we 
go back and forth. When you finally get the money, I get to start the project. We 
want to get out in advance of that. We want to build the interstate highway system. 
We want to be judicious about it, but we want to move towards the model of 
getting a congressional appropriation bill through and starting the project. 

Creating Data Standards 

The limitations of the existing infrastructure for addressing the changing needs of the business were 
exposed by the creation of the Customer Support Center and J&J HCS. When the Customer Support 
Center attempted to sell for national accounts, differing data definitions hindered efforts to understand 
how much total business any one customer did with J&J and what services J&J could offer. Jim Litts 

If you go to a mass merchandiser as Johnson & Johnson, you can walk in there as 
the number one or number two non-food manufacturer on that account. At the 
same time you can bring things like pharmaceutical, professional, and pharmacy 
information and counsel and advice and ideas from our other J&J companies. If 
you do that together as J&J, you have a tremendous ability to start opening doors 
that you cannot do if you're one company selling sanitary protection products. You 
can go in there as J&J and have this story. So the guy says, "Okay, good. Give me 
some help." You turn around and you say, "Good grief, none of this stuff adds 
up." You spent an inordinate amount of time just trying to get the information 
together. Then you take it in there and they ask three questions, and you've got to 
go back and do it all over again. 

Steve Piron, Vice President of Information Architecture, observed that franchise management also 
demanded cross-company information that was not available from existing systems: 

J&J France, for example, defined for itself the information it needed about the 
French Consumer business. But when we moved to worldwide franchises, like our 
shampoo business, we were stuck from an information point of view because we 
had product codes and product costs and definitions around the customer defined 
on a country level, and not a region [e.g. Europe] or a worldwide level. So we had 
apples and oranges from an information point of view. 

Even where companies used common systems, they had, on occasion, abandoned common data 
definitions. The Consumer companies, for example, had all adopted the same homegrown order entry 
system, but as Jim Litts explained, they did not all adopt the data definitions: 

Sales reporting for [Consumer] companies all comes out of the Group order entry 
system and is passed back in a uniform kind of format and information display. 
Then every company takes that and alters it. So when you try to add it back up 
again, or if you take the order entry system numbers and then go down and have 
a conversation with the company, they're different. 

Building the IT L'nit of the Future 

To support J&J's efforts to increase cross-company cooperation and coordination, Ed Parrish, the 
corporation's chief information officer, identified three initiatives intended to enable easy sharing of 
information across companies: (1) standardizing data definitions and formats for key data elements on 
a world-wide basis, (2) defining and establishing the information technology infrastructure needed to share 
data and information electronically, and (3) developing and applying IT expertise as a corporate rather 
than a company function. These internal IT efforts were expected to increase the effectiveness of the IS 
unit and allow more time and attention for strategic applications of information technology. 

Steve Piron was heading up efforts to standardize critical data definitions and the methods for 
communicating them. Along with a team of IT professionals who would recommend data standards to 
higher level IT and business managers, Piron was working to develop a data warehouse accessible, as 
needed, by J&J decision makers. Piron's teams would be putting a process in place that defined standard 
data definitions in critical areas like customer, product, competitor, supplier, and then determine which 
would be shared on a world-wide basis, which was a regional data item, and which was a country data 
item. They would also define processes to see that the standards got implemented in transaction 
processing systems around the world. HCS and the Customer Support Center had already specified some 
definitions and these would be presented to other companies. 

NCS had started to define specific standards for hardware and software such as desktop office suites and 
LAN operating systems. Matthew Martin of HCS noted that these standards would be HCS standards, 
and that this expectation had been communicated to all eighteen HCS companies. Executives at each 
company had been asked to specify needed dollar resources, and time frames in order to "get up to 
speed." Ed Parrish noted that one company that was part of the HCS initiative had been adamantly 
opposed to standards, but quickly moved toward implementing them once HCS had made that 
commitment. Parrish said he would target 80% acceptance of standards, because by that point the other 
20% would stand out and senior management would quickly bring them into line, if appropriate. 

Parrish planned three efforts to position IT as a corporate function. First, he would initiate training 
programs in which IT staff throughout the corporation were taught what they needed to know about IT 
at Johnson & Johnson. Second, he would impact pay and performance by having Group Vice Presidents 
share their performance evaluations of IT directors with each IT director's company president. Finally, 
he would take over succession planning, so that when IT director positions opened up, the company 
president would receive a short list of candidates from which to choose a successor. 


Johnson & Johnson had over one hundred years of experience in decentralized management practices, but 
the company needed to rapidly adopt processes that enabled it to share data across business units and 
practice cross-company cooperation. IT management identified several strategies to accelerate the process 
of implementing Parrish's initiatives: 

• Some managers argued for adopting common systems to help implement new data definitions. 

Others, however, felt that common systems would not meet individual business needs and that 
clear definitions guiding development of translation programs were key to creating a successful 
data warehouse. 

• Outsourcing was suggested as a means for forcing changes that consensus processes would be 
slow to embrace. Practices that involved personnel shifts and adoption of standards might be 
more easily accepted when mandated by external parties. 

• Parrish noted that communicating standards and data definitions to senior management would 
help implementation efforts. As business executives decided they needed new kinds of 
information, they could enlist support for standards and force agreement on data definitions. 

Johnson & Johnson would likely employ all these strategies as it attempted to adapt to dynamic business 


Appendix A 
Representative Sample of J&J Operating Companies 

Cilag — manufactures and markets products primarily discovered and/or developed by the R.W. 
Johnson Pharmaceutical Research Institute, includes products in areas such as fertility control, 
dermatology, and immunoregulatory peptides. Family of operating companies includes Cilag 
G.m.b.H. in Germany, Cilag-Medicamenta Limitada in Portugal, and Janssen-Cilag Pty. Ltd. in 

Ethicon — develops and markets innovative products for surgeons. It produces thousands of 
sutures, ligatures and related products. Family of operating companies includes Ethicon, Inc. in 
the United States, Ethicon S.A. in France, Ethicon Endo-Surgery in Japan and Ethicon Limited 
in Scotland. 

Janssen Pharmaceutica — produces a broad range of pharmaceutical products in areas such as 
allergy, anesthesiology, gastroenterology, psychiatry, and cariovascular disease. Family of 
companies includes Xian-Janssen Pharmaceutical Co. Ltd in China, Janssen Pharmaceutica, 
Limited in South Africa, Janssen Pharmaceutica S.A.C.I. in Greece and Janssen Farmaceutica, 
S.A. de C.V. in Mexico. 

Johnson & Johnson Consumer Products, Inc. — provides wound care, baby care, oral care and 
skin care products. These are manufactured and sold in companies throughout the world, 
including Johnson & Johnson de Venezuela, S.A., Johnson & Johnson Inc. in Canada, Johnson 
& Johnson Limited in Zambia, and Johnson &. Johson/Gaba B.V. in The Netherlands. 

Johnson & Johnson Medical Inc. — provides products for wound management and patient care, 
such as intravenous catheters, disposable surgical packs, latex surgical and medical gloves, and 
wound care sponges and dressings. Family of companies includes Johnson & Johnson Medical 
Thailand, Johnson & Johnson Medical in the Philippines, Johnson & Johnson Medical in Ireland 
and Johnson & Johnson Medical AG in Switzerland. 

Johnson & Johnson Professional Inc. — develops and markets products under the CODMAN 
brand for the surgical treatment of central nervous systems disorders and under the J&J 
Orthopaedics brand for musculoskeletal system repairs. Family of companies includes Johnson 
& Johnson Professional Products Ltd. in England, Johnson & Johnson Professional Products in 
Sweden, and Johnson & Johnson Professional Products, G.m.b.H. in Germany. 

Ortho Diagnostic Systems Inc — provides diagnostic reagent and instrument systems to hospital 
laboratories, commercial clinical laboratories and blood donor centers, such as diagnostic systems 
for coagulation, AIDS, hepatitis and other infectious diseases. Ortho Diagnostics is found in 
Canada, France, Japan, Spain and the United States. 

Vistakon — produces and markets the leading disposable contact lens. This operating company 
is based in the United States. 




























D) CU 


£ (T> 


o c 



£ 3 

a) a. 
2 E 































































"-•— » 












E o 













— E 



W O) 


< g 








91 .£ 




,2 DJ 


CO c 







CO -X 


,« «n w 


2 <D <d 



o "- 1 ■> 
13 D- £: 

«J rr a> 

0) t- CO 




a> w 

•2 o w 





J^ Q. UJ 


J= < 




(3 (D 



JC o 


CO £ 

2 <» 





Johnson & Johnson: 
Building an Infrastructure to Support Global Operations 

Teaching Note 

Jeanne Ross, MIT Center for Information Systems Research 

The Johnson & Johnson case poses the arguments for and against establishing centralized IT 
infrastructure support and IT standards in decentralized, global organizations. It describes changes 
in J&J's business environment that expose the limitations of the firm's traditional structure and 
culture, and then examines the role of information technology to help address those limitations. The 
fundamental problem J&J faces is one that many firms are experiencing — how can the autonomous 
business units of a large firm present a single face to global customers. The case asks students to 
consider not only what IT infrastructure would best address current business demands but also how 
J&J should go about implementing such an infrastructure. 

The case can be used in either IS major or general MBA courses. It includes limited technical 
information on J&J's network design, which will be more readily understood by technically-oriented 
students. We encourage faculty to help less technical students understand this information, because 
we anticipate that business managers will increasingly be involved in decisions on IT infrastructure. 
This means that business managers should be conversant in networking issues, so they can better 
participate in discussions on IT infrastructure as well as understand the organizational and financial 
commitment required to make it effective. 

Recommended Discussion Questions 

1. Describe J&J's organizational structure and culture. How important are this structure and 
culture to the firm's past and continuing success? 

2. Describe the structure of the IT unit that has resulted from J&J's decentralized organization. 
Why is infrastructure support more decentralized in the U.S. than in Europe? 

3. Is the European Network Services model an appropriate model for Bob Chaput's Network 
Services and Computing unit? What alternatives does Chaput have? 

4. What strategies would you suggest to Ed Parrish, the CIO, if he wants to work toward more 
centralized IT management? 

5. What do you think J&J's organizational structure will be five or ten years from now? What 
are the implications for infrastructure development today? 

Suggested Additional Readings 

All of the following readings provide useful insights for discussion of the J&J case. Faculty can 
decide which issues are of greatest interest. 

Broadbent, M. and P. Weill, "Infrastructure Mix and Match," MIS, October, 1994, pp. 52-55. 

Goodhue, D.L., Kirsch, LJ., Quillard, J. A. and Wybo, M.D. "Strategic Data Planning: Lessons from 
the Field." MIS Quarterly, Volume 16, Number 1, March 1992, pp. 11-34. 

Keen, P.G.W. and J.M. Cummins Networks in Action, Wadsworth Publishing Company, Belmont, 
California. 1994, chapters 6 and 7. 

Kotter, J. P. "Leading Change: Why Transformation Efforts Fail," Harvard Business Review, March- 
April 1995, pp. 59-67. 

Lacity, M.C and Hirshheim, R. The Information Systems Outsourcing Bandwagon," Sloan 
Management Review, Volume 35, Number 1, Fall 1993, pp. 73-86. 

Nichols, N.A. "Medicine, Management, and Mergers: An Interview with Merck's P. Roy Vagelos," 
Harvard Business Review, November-December 1994, pp. 104-114. 

Von Simson, E.M. 'The 'Centrally Decentralized' IS Organization," Harvard Business Review, July- 
August 1990, pp. 158-162. 

Class Discussion 

The outline for a class discussion on this case would cover the following topics: 

A. Overview of J&J, business strategy, organizational structure and culture 

B. Discussion of business imperatives 

C. Overview of J&J IT Organization 

1. Implications of decentralized operating companies 

2. Discussion of the benefits and limitations of the European infrastructure model 

D. Discussion of Infrastructure Design for J&J 

1. Identification of alternative designs 

2. Discussion of strategies to move toward greater centralization 

E. J&J's Business Vision and Implications for Infrastructure Development 

J&J Culture and Structure — The recommended class structure would start by discussing J&J's 
traditional structure and culture. The instructor can ask about the costs and benefits of the highly 
decentralized structure and probe as to why J&J management feels the need to tinker with what has 
been a highly successful organization. The class should discuss the potential impacts of the 
"centralizing" initiatives management has undertaken in recent years (e.g. organization of groups and 
creation of franchise managers, listed at the bottom of page 2 and top of page 3). The class should 
observe that J&J's decentralized organizational structures has not only led to a lack of mechanisms 
and processes for exchanging information, it has discouraged individuals from wanting to exchange 

information across operating companies. Thus, J&J employees have neither the technology, nor the 
will, to support information sharing. 

Changes in Business Conditions — As stated throughout the case, J&J is facing new customer 
demands. In particular, large customers, which include retailers, hospitals and health care providers, 
want to deal with fewer individuals and companies at J&J. This is why HCS and the Customer 
Support Center have been formed. The Nichols article discusses changes in the industry in greater 
depth and might be assigned as additional reading. Business conditions make clear that J&J's 
traditional organizational structure and processes will come under fire. 

IT Unit Organization — The discussion of J&J's structure and business conditions should lead into 
a discussion of the IT organization. The autonomous nature of the operating companies means that, 
for the most part, operating companies meet their own computing needs. Moreover, business needs 
have only recently created a need for widespread coordination of activities across operating 
companies. Students should be aware that it is business reasons, not technological limitations, that 
have led to the decentralized organizational structure. 

The discussion should explore the implications of this decentralized approach to developing a 
computing infrastructure. Each operating company has established its own priorities, decided which 
technologies to adopt and which to avoid, purchased or developed solutions to its business problems, 
and defined data to meet its unique business needs. We can assume that the level of sophistication 
of both IT personnel and business users varies considerably across organizations. Similarly, the 
companies differ in the value they attach to IT and the amount they have invested. There are 
pockets of expertise within the corporation for many different technologies, but this expertise has 
typically not been shared across companies. So the corporation has some valuable resources in the 
form of IT expertise and experience, but it is not leveraging those resources across companies. On 
the other hand, because the businesses control their IT resources, they are able to ensure that their 
top priorities are addressed and they can see the results of their IT investments. 

The corporate infrastructure has been minimal. Bob Chaput's organization designed voice 
telecommunications networks for each company and managed the contracts with service providers, 
but the individual operating companies actually worked with the telephone companies to implement 
their systems solutions and administer their systems. On the data side NCS also managed a physical 
network, including bandwidth and physical devices like routers, that enable data transfer between 
operating companies and the data center, but most of NCS supported the data center and mainframe 
processing for the corporation's U.S. companies. Transaction processing for sales, billing, payables 
and other routine processes had been centralized in order to make efficient use of mainframes. 

Students should understand that this centralized transaction processing capability means that many 
operating companies have established electronic linkages with central computers. However, the 
communications environment is highly constrained, using an SNA architecture in which operating 
company computers communicate with the host at corporate. (See Keen and Cummins for a 
discussion of SNA.) The environments within operating companies are characterized by a variety 
of hardware, operating systems, database management systems and application programs. Thus, in 
a pure SNA environment, the companies do not have electronic linkages with one another. 

Benefits and Limitations of European Structure — Although the firm has a mostly decentralized IT 
infrastructure, European operating companies have steadily come to rely on central services. 

Students should observe the business reason why this developed (one large operating company could 
otter resources and expertise that were unavailable within small operating companies). Because 
many of the European businesses sought out the help of the networking experts at Janssen 
(pronounced yon'son), and other IT support from Corporate, Europe has not had the variety in 
computing platforms that the U.S. has had. 

The significance of the European environment is that it is manageable from a central location. 
Because European Network Services has implemented a router-based TCP/IP architecture, data can 
be exchanged among sites as long as data definitions and formats are common. Because essentially 
all European regional servers are connected to the network, the Network Services staff can provide 
centralized services to leverage their expertise and provide low-cost infrastructure support. They 
could establish standard protocols for printing, file transfers, and back-up, and centralize help for 
desktop applications, groupware, and e-mail. Thus, most European operating companies do not 
need networking and infrastructure applications expertise on site. 

It is worthwhile to explore what European companies are giving up in exchange for cost-effective 
infrastructure management. Students should readily recognize that management will experience 
some discomfort with not having local expertise when something goes wrong. Indeed, the long-term 
success of centralized support infrastructures is at least partly dependent upon the ability of the 
centralized unit to identify and fix network problems quickly, to provide help desk service to remote 
sites via telephone and e-mail, and to contract with reliable third parties to service individual 
companies' on-site needs. 

It is also worth noting critical success factors associated with running a centralized unit like 
European Network Services. Clearly, because they were selling their services to other companies, 
they had exhibited expertise and an ability to address the networking needs of other companies at 
a fair price (in other words, the operating companies felt they were receiving good value). It is also 
essential that European Network staff closely monitor equipment and establish standards for 
technology that meet business needs and that they are capable of supporting. The instructor will 
want to establish these and other CSF's for their later discussion of what Bob Chaput might want 
to do and how he might do it. 

In discussing differences between corporate NCS and the European unit, note that corporate IT has 
responsibility for all J&.J networks. The European Network Services team reports into Bob Chaput 
and has taken on responsibility for the networks, and for support of distributed systems, for 
operating companies in Europe and Northern Africa. Chaput notes that he would like to institute 
similar regional centers possibly in Asia and South America, but he would also like to make the 
European center a model for North American, since a large cost, quality, and service opportunity 
exists there. 

Alternative Designs for the Corporate Infrastructure — It is worthwhile to list what J&J must do to 
respond to its changing business conditions. Students should recognize that presenting a single face 
to the customer will demand more than information technology initiatives, but IT will be an 
important enabler. Non-IT initiatives might include efforts such as developing rewards and 
incentives for achieving corporate goals, increasing cross-company transfers, and possibly even 
introducing more umbrella organizations like HCS and the Customer Support Center. IT initiatives 
fall into 2 categories: (1) generating sharable data through standardized data definitions and (2) 
providing electronic linkages and support to facilitate exchange of that data. Students should be 

aware that it is possible to have one of these without the other, and that the European Network 
Services model only addresses the latter. 

J&J is attempting to address the data standardization challenge through Steve Piron's team. The 
class may want to debate the idea of defining data definitions across the categories listed on page 
8 and then categorizing them as world-wide, country or regional. (The Goodhue et al article 
provides interesting background reading on large-scale data modeling efforts, which is relevant to 
this discussion.) It should be clear that some data standardization is necessary, but the only success 
that J&J has had in this area at the time of the case is that which was driven by immediate need — 
the need for HCS and Customer Service Center companies to share customer data. 

Some faculty may want their classes to debate the common systems issue, posed at the end of the 
case. Common systems, either those produced in-house, or systems like SAP, force some 
commonality of practice and consequently some standardization of data. Of course, as noted in the 
case, individual companies may manipulate data from the common systems such that it becomes 
nonstandard. In addition, some managers resist common systems on the basis that they do not 
address the unique needs of their business (although this is becoming somewhat less of a concern 
as newer client-server systems offer increased flexibility). A discussion of common systems should 
note that in the short-term, they may be expensive to implement and they are inevitably slower to 
implement than might be expected. The data standardization process (and, to a large extent, 
workflow standardization) is a necessary precursor to implementation of most common systems. So, 
gradual acceptance of common systems may be more realistic than reliance on common systems to 
introduce data standards. 

Chaput's expressed desire to use the European center as a model for corporate network support 
offers one alternative for the infrastructure design at J&J. Students should observe that there are 
alternative approaches. For example, Chaput plans to extend the regional center concept so that 
a corporate center is not really needed. (Students should observe that mechanisms for coordinating 
across the regional centers would be essential. In fact, they may outweigh any benefits to be realized 
from this model.) Alternatively, given that the U.S. companies are often big and self-sufficient, J&J 
could work toward some centralization of its IT infrastructure, but remain more decentralized than 
Europe. For example, headquarters could provide 24x7 (24 hour, 7 days a week) support of 
networks, but leave local experts at each site. Similarly, help desk personnel could be centralized 
or distributed. (See Broadbent and Weill for a list of infrastructure services that can be centralized 
or decentralized.) 

Implications of Infrastructure Centralization — The reasons why Chaput wants to pursue the European 
model are evident from the discussion of the benefits of this arrangement, but students should note 
that, unlike Europe, centralization will lead to cultural upheaval in U.S. operating companies. 
Students should consider what might be involved in moving to a more centralized environment: 

• much more centralized or consensus decision making 

• likely reduction in the IT staffs of operating companies 

• service that addresses the common good of the corporation rather than focuses on individual 
needs (i.e. possible service level reduction) 

• greater standardization of hardware, software and operating systems 

• higher monthly charges from corporate IT (but lower unit costs) 

• possible physical relocation of staff 

They should conclude that there will be significant resistance to the idea of centralized infrastructure 

Strategies for Centralization — At this point, students should be asked what it will take for Chaput 
to succeed in his efforts to centralize. In particular, they might focus on the initiatives that J&J is 
undertaking and their likely effectiveness. 

Professional Group shared services: led by the Vice President for Professional Group IT, 
five companies were insourcing telecommunications management to NCS, and examining the 
possibility of insourcing distributed systems management. This is a "quick hit" approach to 
implementing a centralized infrastructure. By providing satisfactory service to this group of 
five companies, NCS could demonstrate its competence and win acceptance from other 
operating companies. There is a risk that the companies will not be satisfied and will thus 
deter other companies from centralized support, but this gradual implementation allows 
Chaput's organization to take on projects of manageable size. The 'quick hit' approach is 
a gradual one and students will want to consider whether a strategy of "quick hits" can have 
the desired impact. 

• Implementation of standard technologies: before adopting technology standards, the IT or 
top business people in an operating company must recognize the value of centralized 
support; otherwise they can rely on personal favorites. Ed Parrish notes that 80% acceptance 
is all that is necessary for standards to take effect, but that will require a great deal of senior 
management support. Note that his role becomes one of communicating standards, and the 
reasons for those standards, to business management. Students are likely to suggest that top 
management could mandate the standards. This has been effective in other organizations, 
but J&J management has tended not to adopt a dictatorial style, thus allowing significant 
foot-dragging when benefits of new initiatives are not apparent to local managers. 

• Outsourcing was suggested as a means of moving the firm to more rapid acceptance of 
centralization. The risks and rewards of outsourcing are discussed in the Lacity and 
Hirshheim article. This alternative offers the opportunity for fairly rich debate as to whether 
enforced organizational change is an appropriate objective for outsourcing. 

• Human resource changes for IT personnel: Parrish talks of pursuing training, new pay and 
reward structures (that would surely reward compliance with standards), and succession 
planning. Students should observe that all of these approaches are attempts to win 
compliance through consensus thinking rather than brute enforcement. They might also 
observe that such approaches are highly consistent with J&J culture. 

J&J's Future Business Model — A final question that students should discuss is J&J's long-range 
business model which the infrastructure must be able to support. In fact, top managers at J&J do 
not have a single vision of the corporation in the next five to ten years. Some anticipate that it will 
abandon the operating center concept and consist of just the three Groups or several HCS/Customer 
Service Center types of businesses. Others feel that the operating company concept will remain 
strong but that large numbers of umbrella organizations will serve to coordinate their efforts. Still 
others feel that the operating centers will be supported by sophisticated communications technologies 
that render franchise management and Groups unnecessary. Students can pose their own view of 
the future and, in each case, propose what it suggests for how Bob Chaput should shape his 


In wrapping up, the instructor will want to emphasize how business demands are constantly changing 
and, in the process, requiring organizational changes. Thus, IT infrastructures must also be 
constantly adapting to new business needs. There is not a single model that will, at a point in time, 
address all the needs of all firms. (The Von Simson article presents an overview of the hybrid 
organizational form, which has become the dominant organizational form.) 

The J&J case offers an interesting look at how a successful organizational form can become less 
responsive over time. It also points out how organizational culture acts as an obstacle to 
organizational change. In J&J's case, the culture is so strong, that efforts by IT management to 
bring about a more centralized infrastructure are primarily focused on change strategies that are 
consistent with the existing culture (policies that rely more on developing consensus than on 
mandating change), even though they will move more slowly to the desired form. 

If desired, the case discussion can close with a discussion of the difficulty of changing successful 
organizations. A significant body of literature has noted that change is much easier to accomplish 
when a company is facing a crisis than when it is enjoying success. This is clearly a challenge for 
J&J and most of its operating companies. The instructor may want to close by reviewing the eight 
reasons why change efforts fail as discussed in Kotter's article (which can be assigned reading or 
simply reviewed by the faculty member). All eight provide useful advice for J&J as the firm attempts 
to address changing business conditions and to Parrish and Chaput, in particular, as they work to 
overcome the resistance to an increasingly centralized IT infrastructure. 


Date Due 



3 9080 01439 0741